Protecting the Medical Commons

by Peter Wehrwein

Public Health scholars grapple with the consequences of as system in which more and more people spend more and more money on medical treatments procedures that produce smaller and smaller health returns.

As a nephrologist, Howard Frazier had seen the slow, wrenching decline of patients with end-stage renal disease many times. First the nausea and vomiting, then the mental disorientation so distressing to families and loved ones, and then the last breath. Doctors did their best in those days, "tweaking and poking and shifting things around a little bit so people didn’t die as fast," says Frazier, an emeritus professor of health policy and management at the School of Public Health. But end-stage renal disease was just that: end-stage.

As Director of the Center for Analysis of Health Practice, Professor Howard Frazier was a mentor and "father figure" to a stream of talented young scholars who would recast research on medical care.

Then the dialysis machine was developed, and in 1954, the first successful kidney transplant was performed at Peter Bent Brigham Hospital. In the years to come, nephrology would go from being medicine that cared to medicine that cured. Money, equipment, and medical talent flowed into the specialty. When there was an outcry over a shortage of dialysis machines and treatment in the early 1970s, Congress responded by setting up a special Medicare-type fund just for end-stage renal disease patients. A heady mixture, this combination of money and technology, but for Frazier it also meant a disturbing paradox.

"There I was, enormously relieved by the shifting of an ethical burden–because resources were available, and it wasn’t my responsibility to somehow scrounge and come up with them. On the other hand, it was just a wasteful way of deploying very limited resources. To put someone of 80 on chronic dialysis–which people argued you have to do because the treatment is there–is just not sensible. You can’t even afford three bucks to provide immunization for the kids across Huntington Avenue in the projects, and yet you could afford, $35,000 to $40,000-a-year to maintain someone symptomatically uremic but not dead."

Frazier wasn’t the only one troubled by the cognitive dissonance in the American approach to health: the dazzling prowess of curative medicine and the relative neglect of what might be done to keep people from needing that medicine. It was also on the mind of the School’s new dean, Howard Hiatt. (See also: Howard Haym Hiatt: Dean, 1972-1986)A top-rank molecular biologist and former chief of medicine at Boston’s Beth Israel Hospital, Hiatt was neither ignorant nor dismissive of the post-war accomplishments of American medicine. But he was also aware of the blind spots. Precious little was being done in the way of systematic analysis of medical treatments–what they cost, what the real gains were. And there was the related issue of people who weren’t getting this care.

"Here we were within the network of Harvard’s teaching hospitals, paying extraordinary attention to the patients, but we had very little involvement with people who didn’t come to us–especially people who didn’t get into the system at all," says Hiatt.

Over the next decade or so, the Harvard School of Public Health became the place where the blind spots, paradoxes, and cross-currents of the American health-care system were put under the bright, searching lights of systematic evaluation and examination. In a research effort unprecedented both in terms of scope and method, faculty at the School studied how decisions were made in American medicine, the motivations behind those decisions, and the consequences both for patients’ and the nation’s health.

To do the job, Hiatt unabashedly imported talent from other parts of Harvard and techniques from other disciplines, most notably from statistics, economics, and decision theory. He galvanized the effort by organizing, with statistician Frederick Mosteller’s help, the Seminar in Health and Medicine, a kind of university-wide think-tank of 100 or so professors and clinicians who met regularly for about three years. The School’s Department of Public Health Administration, renamed the Department of Health Policy and Management, was transformed into a hot-bed of original research and scholarship. Members of the seminar and department faculty wrote prolifically and played a major part in shaping a body of research about the cost-effectiveness of American health-care. Some of the landmark books included Costs, Risks and Benefits of Surgery; Hypertension: A Policy Perspective; and Clinical Decision Analysis. Franz Ingelfinger, the editor of The New England Journal of Medicine was receptive to this new, quantitative approach to medicine and provided Hiatt, Professor Milton Weinstein, and Harvey V. Fineberg, then a young health policy professor, a high- profile forum for their ideas.

The School’s health policy researchers also made waves outside of academic circles. Professor William Hsiao’s Resource-Based Relative Value Scale (rbrvs) changed how Medicare pays doctors and underscored the need for more primary-care physicians. The School’s extensive study of medical errors in New York State hospitals made national headlines in 1990 and injected reliable data into the nation’s fractious debate over medical malpractice. Professor Robert Blendon, the Roger Irving Lee Professor of Health Policy and Management, emerged as the nation’s foremost expert on the nuances of American attitudes about health care and medicine during the prolonged debate over health-care reform during the first two years of the Clinton administration.

As the Sturm and Drang over the nation’s medical costs mounted in the 1970s, other schools and institutions–most notably the Rand Corporation–also started to study and measure the workings of the American health-care system. Frazier says the School set the pace. "I don’t think there was a single place with the same intensity and scale of effort," he says. Hsiao credits Hiatt and his vision of rigorous study of health care and medicine with "revolutionizing the field of public health."

Historical precedent did exist for evaluating medicine. Ernest Codman, an iconoclastic surgeon at Massachusetts General Hospital, was ahead of his time in the 1910s when he argued for systematic data collection and analysis on surgeons. About a decade later, the privately-funded Committee on the Costs of Medical Care generated for the first time some statistics now taken for granted: total national expenditure on health-care ($3.6 billion in 1929); a breakdown of how those expenditures were divided between physicians and hospitals; and some telling numbers on how medical expenditures varied with income. Roger Irving Lee, the Boston physician and Harvard Corporation member so influential in founding the School, was a member of that committee and a prominent advocate of applying economic thinking to health and medicine. In the 1950s, Paul Lembcke, a surgeon at Johns Hopkins, proposed the use of "medical audits" as a way of comparing the performances of physicians and hospitals. At about the same time, Henry K. Beecher, an anesthesiologist at Massachusetts General Hospital, asked a young Harvard statistician named Frederick Mosteller to help him and members of his department design and analyze studies of anesthesia and analgesics.

After leading an interdisciplinary seminar that sowed the seeds for medical care research at the School, Frederick Mosteller would go on to head the Department of Health Policy and Management and the Deaprtment of Biostatistics.

Starting in about 1970, however, several forces came to bear to give these questions about the cost and quality of medicine a new edge and relevance. The nation’s medical bill had skyrocketed, with total health expenditures almost tripling between 1960, when $27.1 billion was spent, and 1970, when the figure reached $74.4 billion. The creation of the Medicare and Medicaid programs in 1965 meant that taxpayers were footing a growing chunk of that bill. When the Medicare-style program for kidney dialysis patients was created in 1972, Congress was told it would cost $200 million. Four years later, the actual costs were double that amount. Dialysis became a case in point for what might happen to the nation’s balance sheet if technologically sophisticated, medically effective, but expensive clinical care became widely available.

Yet Americans were in a "love" phase of the country’s love-hate relationship with medical technology. Heart transplants and coronary bypasses were dazzling treatments for the nation’s number one killer. President Richard Nixon’s 1971 announcement of a war against cancer promised to deliver magic bullets to slay number two. If Americans were enthusiastic about medicine’s technical virtuosity, it was partly because they weren’t seeing the bills, at least not directly. World War II-era tax laws that made insurance premiums tax-deductible allowed a system of employer-based health insurance to flourish. Fewer and fewer patients had to pay for their health care out of their own pocket.

Hiatt was among a relative few who saw a darker side to the medical spending spree on medical care. In his 1975 monograph, "Protecting the Medical Commons: Who is Responsible?" published in The New England Journal of Medicine, Hiatt analogized the American appetite for medical care to Garrett Hardin’s now-famous parable about herdsmen ruining their common pasture by pursuing their self-interest. "Surely nobody would quarrel with the proposition that there is a limit to the resources any society can devote to medical care," wrote Hiatt, "and few would question the suggestion that we are approaching such a limit."

"I honestly think that Howard envisioned that eventually the rise in medical costs and technology would be unsustainable," says Weinstein. "[But] the political and social currents were all going in the opposite direction; that is, more technology is better and medical care is ‘free’ because someone else pays for it."

In taking this somewhat contrarian approach, Hiatt had to basically start from scratch–and without spending a lot of money. He settled on the idea of an interdisciplinary seminar and selected Mosteller to lead it. A few years earlier, Mosteller had played a key role in a famous Harvard seminar on education. With his experience in medicine and delight in solving statistical research problems, Mosteller was doubly qualified to lead a similar kind of seminar at the School. In the fall of 1972, the Seminar in Health and Medicine held its first meeting. "It wasn’t very grand," recalls Hiatt, "because we didn’t want people coming for a steak dinner. We wanted them to come to work."

Mosteller says the first couple of meetings there were some grumblings from fellow professors about what the seminar would do exactly. "I said the point was to do what professors do, which is to profess, write papers and books," says Mosteller. When it convened as a whole, the legislature-sized seminar met in Snyder Auditorium, but much of the real work was done by smaller groups organized around topics. Formally, the seminar lasted about three years, but some of the smaller groups have kept on meeting. (Mosteller’s surgery group still meets on occasion.) Among the slew of papers and books produced by seminar participants that laid the groundwork for what is now called "evidence-based" medicine was the landmark 1977 book, Costs, Risks and Benefits of Surgery. Co-edited by Mosteller, John P. Bunker (who chaired the seminar its second year), and Benjamin A. Barnes with chapters written by other members of the surgery group, the book laid out the main ideas of cost-benefit, decision analysis and applied them to a range of surgical interventions, including coronary bypass, cholecystectomy for silent gallstones, radical mastectomy for breast cancer, and hysterectomies.

Frazier credits Mosteller for the seminar’s success: "He is the only person I know who can enter a room on one side and leave it on the other, having gotten commitments to write four chapters of a book and made arrangements for two collaborations!"

The other organizational device used by Hiatt was the extra-departmental center. Soon after becoming dean, Hiatt launched the Center for the Evaluation of Clinical Procedures (CECLIP). At least in the beginning, "center" suggested something a good deal more grand than the underlying reality. Weinstein remembers sharing a basement office in the School’s old main building at 55 Shattuck Street with Peter Braun, now a lecturer at the School. "We were the Center for Evaluation of Clinical Procedures!" he says. CECLIP was later renamed the Center for the Analysis of Health Practices, and Frazier, who was on Hiatt’s staff when he was chief of medicine at Beth Israel Hospital in Boston, was recruited to lead it, a position he held for eight years.

Professor Milton Weinstein, one of the early members of CAHP, would pioneer cost-effectiveness research on medical care.

Weinstein says Frazier was an effective "father figure" to the young investigators at CAHP; Hiatt provided the intellectual spark. Mosteller says Frazier’s practical medical know-how was essential to the successful grant application that kept the CAHP going. He also advised the cadre of young investigators on their research projects and papers, says Mosteller, though on a behind-the-scenes basis because he rarely signed on as co-author. "Howard put in a lot of time improving other people’s work," says Mosteller. Fineberg, a member of the Center from 1973-1978, says it was an exceptionally close-knit and talented group of researchers who continually batted around new ideas, testing and challenging them.

Gradually, the approaches and some of the key personalities involved in these satellite operations were folded into the main organization of the School. Mosteller again was a key figure, serving as chair of the biostatistics department for several years in the late 1970s and recruiting Marvin Zelen from the State University of New York at Buffalo. Zelen would turn the biostat department into a world leader in randomized clinical trials, one of the mainstays of cost-effective, evidence-based medicine. Mosteller chaired the Department of Health Policy and Management from 1981 to 1987, steering it toward an emphasis on research and original scholarship.

One of the responses to the growing sense of an eroding commons and medical spending reaching its limits was the development of cost-effectiveness research. In vastly over-simplified terms, cost effectiveness research was a way of finding the biggest "bang" in health gain for each "buck" of health expenditure. Weinstein, in a seminal 1977 article in the New England Journal of Medicine, co-authored with William Stason, put it this way: "...the underlying premise of cost-effectiveness in health problems is that for any given level of resources available, society wishes to maximize the total aggregate health benefits conferred." A typical problem might be comparing how many years of life are gained by coronary bypass operations to a program aimed at lowering everyone’s cholesterol levels.

Weinstein says cost-effectiveness health research involves wedding epidemiology to clinical information: Epidemiology gives researchers a handle on the size of the health problem they are dealing with, while clinical information is used to factor in the effectiveness of the various treatments available. In doing cost-effectiveness analyses, the various nuances of "health" also have to be considered. Is the goal just to reduce mortality and have people live longer? Or does health mean being able to run up the stairs? Weinstein and other cost-effectiveness experts have worked at perfecting a comprehensive, quantitative measurement of health outcomes. Weinstein has also developed elaborate questionnaires to gauge and summarize the trade-offs people make when considering quality of life. So, for example, Weinstein might ask, "How much severe nausea are people willing to endure for an extra year of life?"

Calculating the costs involved is no less complicated. Tallying up the direct medical costs of the treatment and the costs associated with any side effects is relatively straightforward. But then the costs of treating the more serious illness that was avoided by early treatment has to be subtracted. Now add back in the costs of treating diseases that wouldn’t have had a chance to occur–say cancer–if the patient had died from the earlier condition. Because there is usually a lag between cost and benefit in medicine (especially when chronic illnesses are involved) all these cost calculations must be discounted to take into account the lower value of future dollars.

Weinstein’s 1976 book, Hypertension: A Policy Perspective, co-authored with Stason, was the full-blown debut of cost-effectiveness research. In it, Weinstein and Stason published their findings that community screening programs were not a cost-effective way for dealing with hypertension. Instead, their numbers showed that targeted approaches aimed at high-incidence populations were more cost-effective. This was the first of several instances in which Weinstein’s research challenged conventional public health wisdom. In 1991, Weinstein published research showing that the most cost-effective cholesterol-lowering strategy should be aimed at treating heart attack victims, not folks with high cholesterol readings who hadn’t had a heart attack. His paper induced a government panel to put a new emphasis on so-called secondary prevention. And this February, Weinstein was co-author of a study published in the Journal of the American Medical Association that suggested that the marked decline in heart disease mortality in the 1980s was due in surprisingly small part to healthier lifestyles. Rather, Weinstein and his co-investigators found that the mortality decline was mainly attributable to advances in the medical treatment for heart disease, including drugs and bypass surgery. "That is heresy at a school of public health," says Weinstein, "but that is what’s happening, we think, based on the data we have and the model we built."

Hsiao, a newcomer to academe after working in Washington, D.C., and in the insurance industry as an actuary, admired the brilliance of the kind of cost-effectiveness studies that Weinstein and Stason were doing. Yet he was troubled by what he saw as the field’s underlying idealism. "I gradually appreciated that cost-effectiveness analysis has a fundamental underlying presumption, namely, that the truth shall set you free," says Hsiao with a laugh. "My perception of human beings is that truth matters, but so do incentives–money and risk. That means how you pay doctors and hospitals may have a greater influence on how they practice medicine." And American medicine was full of distorted incentives that, for example, paid doctors handsomely for reading an ekg, which can be done in a couple of minutes, but paid poorly for carefully examining and diagnosing a new patient complaining about a fever and muscle pains.

"You have to reward physicains for doing the right thing," says Economist William Hsaio, whose Resource Based Relative Value Scale seeks to compensate primary care doctors.

In 1977, Hsiao co-authored a paper with Stason titled "Rational Price for Physician Services" that sketched out a framework for a more rational schedule for physician payment that Hsiao called a relative value scale. "It was like a pebble dropping into the ocean," Hsiao told Robin Marantz Henig, author of The People’s Health, a book about the recent history of the School. "It was read by my mother and my wife. Neither one understood it." But by 1986, the ripples from Hsiao’s had become waves as his system was about to be used to adjust Medicare payment schedules. Almost $7 million in special Congressional funding had turned Hsiao’s little-noticed academic efforts into a busy, policy-making enterprise employing 10 faculty members (mostly on a part-time basis) and a similar number of full-time researchers.

Three years later, Hsiao’ brainchild, renamed the Resource-Based Relative Value Scale, or RBRVS, became the system for Medicare payment. The main thrust of Hsiao’s plan was to shift Medicare payments away from specialty services and specialists to primary care and the doctors who provide it. Hsiao says the full effect of his plan was blunted by changes made by Congress under lobbying pressure from medical specialists. Even so, Hsiao figures the annual savings due to rbrvs runs in the range of 10 to 50 times the government’s $7 million investment in its development.

The effect of Hsiao’s payment scheme is now being felt throughout the medical system as insurance companies and managed care organizations use rbrvs, or modified versions of it, to set payment schedules for their member physicians. Hsiao also gets credit for being among the first to effectively highlight the surplus of specialists in American medicine. "If you look now, there is no more debate about having too many specialists," notes Hsiao. " The question now is how do we close down the training slots."

Hsiao’s focus is now overseas as he advises governments in China, Taiwan, and several other countries dealing with health-sector reform as booming economic growth and a shift toward market economics erode existing, government-run systems. Cultural values and professional ethics notwithstanding, incentives and risk avoidance have the same pull on doctors and hospitals everywhere, Hsiao says. In China, for example, too many patients are given intravenous fluids for no good medical reason. Why? Because, says Hsiao, doctors there are paid something extra if the patient is hooked up to an IV line.

Hsiao rejects the notion that he has a cynical view of doctors. "I wouldn’t say it is cynicism. I would say it is realism. By acknowledging that physicians are human, rather than believing that they somehow rise above all temptation, we deal with health care in a much more efficient and rational way," he says

"You have to reward physicians for doing the right thing. When you reward them for doing the wrong thing, you are pushing a stone up the hill."

Financial incentives are not the only reason things sometimes go awry in health care. As in any endeavor, people make mistakes through inattention, oversight, and outright negligence. In health care, however, the stakes are higher than in most fields. Slip ups can cause serious harm or even death.

Studies of mistakes and mishaps in medicine had been done, most notably by Don Harper Mills in California. But in the 1980s, researchers at the School launched an investigation of "adverse events" in hospitals that set a new, higher standard for malpractice research. A surge in the price of malpractice insurance premiums paid by doctors was the jumping off point for a more general interest in reforming the malpractice system. Once again taking the broad, university-wide approach he used in other areas, Hiatt teamed up with James Vorenberg, then the dean of Harvard Law School, to form the Harvard Medical Practice Study Group. But the group soon was ensnared in sharp differences of opinion about what exactly the malpractice problem was and how it might be mended. Doctors and other health professionals saw a need for capping damage awards and alternative, less litigious ways of settling claims. Lawyers wanted stiffer sanctions against bad doctors and easier access to the courts. The division existed, in part, because neither side had much in the way of firm facts. "[We] were mired in these controversies because we suffered from the same information gap that was afflicting legislators and courts," says the preface to A Measure of Malpractice, a 1993 book about the School’s malpractice research co-authored by Hiatt and some of his co-investigators.

To fill the gap, Hiatt and others decided to do a study aimed at finding out how much medical injury occurred, the seriousness of those injuries, and who among the injured sued. When they proposed doing their study in Massachusetts, the researchers got a frosty response: doctors and hospitals were worried the study would waylay pending legislation that would cap pain and suffering awards. Then Hiatt met David Axelrod, the New York State Health Commissioner. Axelrod, a powerful health official who was pushing for improvements in New York hospital care on a number of fronts, embraced the project and persuaded the New York state legislature to spend several million dollars on it. "[The] most important resource that New York offered us was David Axelrod," says A Measure of Malpractice.

Besides Hiatt, the Harvard Malpractice Study research team included Professor of Law and Public Health Troy Brennan; Biostatistics Department Chair Nan Laird; Adjunct Professor of Health Policy Lucian Leape; and Howard Frazier. Hsiao assisted during the planning phase. Paul Weiler, a professor at the law school, and Benjamin Barnes, an associate professor of surgery emeritus at Harvard Medical School, also played key roles. The study had the kind of elegant yet sturdy design that pleases epidemiologists at an aesthetic as well as scientific level. The investigators randomly selected the hospitals to be studied. If they had depended on volunteers, the results might have been biased by hospital administrators avoiding the researchers, fearing that they might discover high negligence rates in their institutions. The researchers also went to great lengths to make sure that their review of hospital records was as objective as possible, training a group of medical records analysts especially for their project. Using a set criteria of 18 clinical "events," the analysts made an initial determination about whether the hospital record suggested that an "adverse event" might have occurred. Those suspicious records were, in turn, looked at by two physician-reviewers. To limit the chance of the physician-reviewers’ bias creeping into the study, the researchers developed the Adverse Event Analysis Form to lead the doctors on a step-by-step review of the record. Leape says one of the key innovations of the study was to recruit local physicians to review the records, which both saved money and took advantage of physicians’ expertise.

The reviewing of the records began in earnest during the summer of 1988. By the time the study was finished, the researchers had sifted through some 30,000 patient records from 51 hospitals. The dragnet found 1,278 "adverse events," about one-fourth of which were identified as injuries caused by provider negligence. Put another way, the study suggested that 4 out of every 100 New Yorkers hospitalized in 1984 were injured during their hospital stay and one-quarter of those injuries was the result of negligence.

The researchers then took the next step to see how this epidemiology of medical injury played out in the legal system. What they found was a malpractice system badly out of sync. On the one hand, there were a large number of "false positives"–malpractice claims against innocent doctors. In fact, most of the malpractice claims filed with New York state officials were not based on any real injury. Fewer still involved negligence. On the other hand, the researchers found an even bigger problem with "false negatives"–patients who had been harmed by negligent care who hadn’t filed any kind of claim of malpractice. The researchers calculated that just 1 in 50 cases of negligent medical care resulted in a malpractice claim . "[We] found several times as many seriously disabled patients who received no legal redress for their injury as innocent doctors who bore the burden of defending against unwarranted malpractice," they wrote in A Measure of Malpractice.

When the results of the malpractice study were first published in 1990, they made front page news in The New York Times. The study spawned a new round of debate about the malpractice system, as well as a stream of like-minded research. The School’s researchers argued in medical journal editorials and elsewhere that their findings pointed to the need for a new, no-fault system for dealing with malpractice that would more speedily and reliably compensate victims. But the combination of a leveling off of malpractice premium prices, which cooled physicians’ interest in the subject, and Axelrod’s disabling stroke in 1991 slowed the political momentum for legislated malpractice reform.

"We were getting there," says Professor Lucian Leape of his efforts to encourage a more constructive response to medical errors.

Even so, Leape says physicians, health-care officials and hospital administrators are beginning to show an interest in reducing medical errors. For one thing, studies like the one Leape published in the Journal of the American Medical Association earlier this year provide evidence that cutting down on errors can save a hospital money. In the jama article, Leape calculated that a 700-bed hospital could save $2.8 million a year by setting up systems to avoid preventable adverse drug events. In the jama article, Leape explained that the Harvard Medical Practice Study found that drugs were the leading cause of adverse hospital-related events. On a national scale, Leape said drug-related morbidity and mortality cost an estimated $76.6 billion. By comparison, he noted in the jama paper, the cost of all diabetes care has been estimated at $45.2 billion.

Last October, Leape was the key-note speaker at a conference on health care error prevention in Rancho Mirage, California, attended by many vips in the health-care and hospital industry. Leape was heartened by what he heard there: "We are getting there. People are now saying that this is a problem we have to deal with."

Evaluation of health care now seems as much a part of the School’s mission, and public health more generally, as research on infectious disease and nutrition. It is easy to forget just how novel systematic study of health-care and medicine was 25 years ago. Weinstein remembers that shortly after he came to the School in 1973, a prominent member of the powerful epidemiology department sarcastically asked him to "say something to me in policy analysis." Hsiao says the attitude was so chilly when he first came to the School that for years he thought he must have been wearing the wrong deodorant.

Times have changed. The Department of Health Policy and Management now boasts 26 primary and 10 secondary faculty members, and among students, the HP&M classes are consistently the most popular. More fundamentally, the School has helped bring what Dean Fineberg has called an "ethic of evaluation" to medicine and health-care more generally. In a 1992 Institute of Medicine report, Fineberg defined the ethic of evaluation as "widespread commitment and expectation among physicians, health care institutions, and the public that medical practices must be evaluated for effectiveness, safety and cost."

"Adopting a systematic, pervasive strategy for evaluation will require public and private resources," Fineberg continued, "but the return in terms of extended lives, reduced suffering and cost savings will be substantial."